OLMS (Open Live Mixing System based on Ardour engine https://openlivemixingsystem.org/ ) Proof Of Concept is running stable. Latest tests on Focusrite Scarlett Solo confirm 3.350ms real-world latency (hardware loopback) with only 1.650ms USB/converter overhead against theoretical buffer size. This isn’t a weekend experiment—it’s a professional live mixing infrastructure: headless, deterministic, RT-optimized.
The Gap We’re Filling Hardware digital mixers: expensive, closed, fixed capabilities. Standard DAWs: not designed for headless reliability in critical live contexts. OLMS bridges this: a distributed, modular mixing engine optimized for RT kernel, turning commodity hardware into professional-grade low-latency audio processing.
What We Need (Specifically) We’re assembling the founding contributor core (3-5 people) to scale the architecture over the next 4 weeks. Requirements:
System competence: Daily Arch Linux user, RT kernel experience, advanced bash scripting, comfortable with JACK2/Ardour stack.
Hardware: USB Class Compliant audio interface to replicate tests across different environments.
Time commitment: 4-6 hours/week for 4 weeks.
NOT looking for: Passive testers, people who “want to learn”, or those expecting plug-and-play software. This requires getting your hands dirty in code and system tuning.
What You Get
Co-Contributor status: Real ownership. You’re not helping—you’re building.
Technical impact: Direct influence on architecture decisions (Xvfb strategy, master/slave topology, routing patterns).
Permanent credits: Your name in the core system and visibility on the official platform.
Direct access: 1:1 collaboration with founder on milestone definition.
How to Join We are looking for experts to define the pro-audio Linux standard. If you are ready to contribute, CONTACT US follow these steps:
If you understand why we’re isolating cores via taskset, how to manage IRQ affinity for USB controllers, and why DMA latency=0 is critical to avoid xruns under load—you’re exactly who we’re looking for.
I haven’t seen reference made to monetization in your posts here, but I do see reference on your website to monetization, including establishment of a proprietary marketplace, offer of potentially paid roles, equity stakes in a company, etc.
I’ve had less than 5ms latency in the studio for years. 32 channels via a PCIe interface. Months of uptime. Well over a dozen years of exclusively using Ardour.
For live use, the cheap XR18 I had and the slightly less cheap X32 I now use have provided me with a low enough latency for my requirements. Both offer similar routing functionality to what I had with SAS Audio routers that I maintained and operated in professional broadcasting facilities, at 1/100th of the price. The X32 lets me record multitrack to an sdcard, so I can leave the much more expensive computer at home.
Assuming your monetization plans work out, what percentage of your profits will you direct back to the projects that yours is completely reliant upon? How will Ardour benefit from your work?
Hi @Dan_Easley , first of all, thank you. I truly appreciate the reality check and the time you spent sharing your experience.
I’ll be very honest: I am a learner. I’m a hobbyist who started this as a personal challenge, collaborating with AI to glue together the tools I love. I am in ‘Phase Zero’ and I have no intention of teaching experts like you how to achieve low latency—I’m here to learn.
To clarify your points:
You’re right, the X32 is a beast. But OLMS isn’t trying to be a better X32; it’s an attempt to create a ‘Software Defined Mixer’. The goal is a headless, modular system where the ‘brain’ isn’t tied to proprietary hardware. It’s for those who want to use any LV2/VST plugin in a live environment with the stability of a dedicated appliance, managed via Web UI.
OLMS is and will remain GPL3. The talk about ‘monetization’ or a ‘marketplace’ on the website is a long-term (perhaps naive) vision for sustainability, but the core engine will always be open.
This project stands entirely on the shoulders of giants like Ardour. Giving back isn’t just an option; it’s a duty. If this ever moves beyond a hobby, a formal contribution plan for upstream projects (donations and code) is a priority. Any bug fixes or optimizations found while stress-testing the headless engine will be pushed back to the community.
I’m just a guy trying to build a tool I’d love to use. I’d be honored to have someone with your ‘dozen years of Ardour’ keep an eye on us and call out my mistakes.
Are we gonna address the elephant in the room that your post and response looks suspiciously AI-generated, just like your logo, and your website?
Are you trying to vibe-code a live mixing system?
Hello. Yes I’m. And yes I use a lot IA as I ever told and documented everywhere all along my threads here. And yes, I’m vibe coding it till I will find other visionary people more skilled than me. I truly appreciate if you can judge the result instead of the form. I hope we can collaborate or at least you try my POC. All the best and than you for being on my post.